Apertis Model Picker

Apertis Model Picker

Opinionated recommendations by task type. Model IDs evolve quickly — always verify current IDs and pricing at https://apertis.ai/pricing?utm_source=apertis-skills&utm_medium=skill-doc&utm_campaign=ecosystem.

Coding & Debugging

Recommended: Claude Sonnet (claude-sonnet-4-6 or latest claude-sonnet-4-*)

• Best coding ability per dollar on the platform
• Excels at multi-file edits, debugging, code review, architecture planning
• Context: 200K tokens

client.chat.completions.create(
    model="claude-sonnet-4-6",
    messages=[{"role": "user", "content": "Refactor this function for better performance..."}]
)

Faster/cheaper: Claude Haiku (claude-haiku-4-5 or latest claude-haiku-4-*) Most capable: Claude Opus (claude-opus-4-*)

Long Context — Large Files, Full Codebases

Recommended: Gemini Flash (gemini-3-flash-preview or latest gemini-3-*)

• 1M+ token context window (fits entire codebases)
• Fast and cost-efficient
• Ideal for: analyzing large repos, summarizing long docs, multi-file reasoning

client.chat.completions.create(
    model="gemini-3-flash-preview",
    messages=[{"role": "user", "content": "Analyze this entire codebase and identify architectural issues..."}]
)

Higher quality: Gemini Pro (gemini-3-pro or gemini-2.5-pro)

Fast Chat — Customer Support, Simple Q&A

Recommended: GPT-4o or GPT-5 mini

• Fast, cost-efficient, reliable structured output
• Good at following system prompt instructions precisely

client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "system", "content": "You are a helpful customer support agent."},
        {"role": "user", "content": "How do I reset my password?"}
    ]
)

Budget alternative: glm-4.5-flash or minimax-m1 — strong multilingual, very low cost

Web Search — Real-Time Information

Recommended: any model + :web suffix

• Works with any non-free model on Apertis
• Returns web_sources[] with title, URL, snippet

response = client.chat.completions.create(
    model="gpt-4o:web",
    messages=[{"role": "user", "content": "What's the latest Claude model?"}]
)
sources = response.choices[0].message.web_sources

Start cheap (gpt-4o:web), upgrade to Claude Sonnet or Gemini Pro if you need deeper synthesis.

Reasoning — Math, Logic, Complex Problem Solving

Recommended: deepseek-r1

• Chain-of-thought reasoning, shows its work
• Strong math and logic performance
• Cost-efficient vs OpenAI o-series

client.chat.completions.create(
    model="deepseek-r1",
    messages=[{"role": "user", "content": "Solve this step by step..."}]
)

OpenAI option: o4-mini (reliable, fast reasoning)

Image Understanding & Vision

Recommended: gpt-4o

• Strong multimodal understanding
• Handles: screenshots, diagrams, charts, handwriting, UI mockups

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{
        "role": "user",
        "content": [
            {"type": "text", "text": "What's in this image?"},
            {"type": "image_url", "image_url": {"url": "data:image/jpeg;base64,..."}}
        ]
    }]
)

Also strong: Gemini Flash/Pro (vision + very long context)

Cost Optimization — Budget Projects, High Volume

Recommended: Gemini Flash or GLM

• gemini-3-flash-preview — very low cost, large context
• glm-4.5-air — multilingual, competitive pricing
• deepseek-v3 — excellent for coding-heavy workloads

Quick Reference

Check live model IDs and pricing at https://apertis.ai/pricing?utm_source=apertis-skills&utm_medium=skill-doc&utm_campaign=ecosystem

Dynamic Model Selection (Recommended for Agents)

Instead of relying on this static list, call the Apertis Recommend API to get the best model for your task at the current time, with live pricing:

curl "https://api.apertis.ai/v1/recommend?task=coding&budget=medium" \
  -H "Authorization: Bearer $APERTIS_API_KEY"

Task types: coding, long-context, fast-chat, reasoning, vision Budget tiers: low (cheapest), medium (balanced), high (best quality)

Example response:

{
  "model": "claude-sonnet-4-6",
  "task": "coding",
  "budget": "medium",
  "input_price_per_1m": 2.40,
  "output_price_per_1m": 12.00,
  "reason": "Best coding ability per dollar. 200K context.",
  "alternatives": [
    { "model": "deepseek-v3", "input_price_per_1m": 0.30, "note": "3x cheaper, good for simpler coding tasks" }
  ]
}

Use the returned model value directly in your API calls. This endpoint always reflects current pricing and model availability.